arXiv cs.CL· Ghazal Khalighinejad, Raghuveer Thirukovalluru, Alexander H. Oh, Bhuwan Dhingra·· 3 天前AI 评分31
研究：文档即图像表示在科学检索中表现不佳，ArXivDoc 基准发布
Document-as-Image Representations Fall Short for Scientific Retrieval
AI 导读
一项研究提出 ArXivDoc 基准，基于科学论文的 LaTeX 源码构建，用于比较文本、图像与多模态表示在科学文档检索中的效果。结果显示，文档即图像表示持续表现次优，且随文档长度增加更明显；纯文本表示最有效，即使面对图表类查询也能借助图注和上下文取得优势；交错文本+图像表示无需专门训练即可超越文档即图像方法。
正文 · 原文
Abstract:Many recent document embedding models are trained on document-as-image representations, embedding rendered pages as images rather than the underlying source. Meanwhile, existing benchmarks for scientific document retrieval, such as ArXivQA and ViDoRe, treat documents as images of pages, implicitly favoring such representations. In this work, we argue that this paradigm is not well-suited for text-rich multimodal scientific documents, where critical evidence is distributed across structured sources, including text, tables, and figures. To study this setting, we introduce ArXivDoc, a new benchmark constructed from the underlying LaTeX sources of scientific papers. Unlike PDF or image-based representations, LaTeX provides direct access to structured elements (e.g., sections, tables, figures, equations), enabling controlled query construction grounded in specific evidence types. We systematically compare text-only, image-based, and multimodal representations across both single-vector and multi-vector retrieval models. Our results show that: (1) document-as-image representations are consistently suboptimal, especially as document length increases; (2) text-based representations are most effective, even for figure-based queries, by leveraging captions and surrounding context; and (3) interleaved text+image representations outperform document-as-image approaches without requiring specialized training.
| Subjects: | Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2604.18508 [cs.IR] |
| (or arXiv:2604.18508v2 [cs.IR] for this version) | |
| https://doi.org/10.48550/arXiv.2604.18508 arXiv-issued DOI via DataCite |
Submission history
From: Ghazal Khalighinejad [view email]
[v1]
Mon, 20 Apr 2026 17:00:17 UTC (328 KB)
[v2]
Tue, 13 Oct 2026 20:40:04 UTC (330 KB)
来源：arXiv cs.CL · arxiv.org